The TASKING LAPACK Performance Libraries

August 1, 2018

TASKING LAPACK Libraries for TriCore/AURIX

Here is our fourth in our series of seven webi­na­rs. This prod­uct is dif­fer­ent from our other tools: this is a library. The LAPACK Per­for­mance Libraries will enable you to call on its proven lin­ear alge­bra pro­cess­ing capa­bil­i­ties to return quick and accu­rate solu­tions for your needs. With sen­sor pro­cess­ing and ADAS, more and more math­e­mat­i­cal com­pu­ta­tions are being required…this Library can help you.

The Lin­ear Alge­bra PACK­age (LAPACK), using Basic Lin­ear Alge­bra Sub­rou­tines (BLAS) per­forms com­plex lin­ear alge­bra manip­u­la­tions on matri­ces of simul­ta­ne­ous equa­tions. If you have some matrix alge­bra stud­ies in your back­ground, I’d guess it was not the most enjoy­able sub­ject. In my class, a stu­dent asked about a home­work prob­lem. The pro­fes­sor looked at the prob­lem, smiled, and said “Wow! That’s a 2 beer prob­lem!” And, indeed, most matrix alge­bra prob­lems ARE 2 (or more) beer prob­lems. And that is where LAPACK excels.

Inter­est­ing­ly, LAPACK is writ­ten in FORTRAN and traces its his­to­ry back to 1992. This long his­to­ry proves the use­ful­ness and accu­ra­cy of the library. For a more detailed look at the work­ings of LAPACK, go HERE.

While there are ver­sions of LAPACK for many proces­sors, TASKING offers the only LAPACK library in C for the Infi­neon AURIX proces­sors.

Click here to view the TASKING LAPACK Library Webi­nar. And, be sure to down­load an eval­u­a­tion to try with your own code at https://www.tasking.com/trial.